Glioblastoma multiforme (GBM) is the most common primary intrinsic brain tumor in adults. Despite maximal therapy, GBM is recurrent and usually fatal in 12-15 months, indicating an urgent need for novel therapies. No single mechanism provides for the failure of current GBM therapies, but several laboratories including our own have shown that GBMs contain cancer cells with stem cell-like characteristics. These glioma stem cells (GSCs) express stem cell markers and can give rise to more differentiated progeny of multiple lineages similar to neural progenitors. Unlike normal neural progenitors, low numbers of GSCs can form tumors in animal models whereas the more differentiated non-stem glioma progeny cannot. GSCs may drive cancer growth by resisting therapy and creating blood vessels to feed the growing tumors. Together these data indicate targeting GSC maintenance, in part through the promotion of differentiation, may have benefit as a patient therapy. I have therefore sought to define mechanisms which may contribute to GSC maintenance by inhibiting differentiation. To address this issue, I focused on inhibitors of bone morphogenetic protein signaling (BMP), a pathway previously shown to promote GSC differentiation and inhibit tumorigenesis. I have identified a BMP inhibitor which may promote GSC maintenance and could be therefore be targeted to provide or enhance pro-differentiation based therapies. Interestingly, this BMP inhibitor may be supplied by the non- stem glioma cells, suggesting that the differentiated progeny of GSCs can provide important instructive cues to GSCs for their maintenance. By investigating these possibilities, I will develop a greater understanding of the interaction of GSCs with other cells in the tumor microenvironment and define novel therapeutics for anti-GSC based therapies.